Hello! I believe that DataStreamer has limited failover support, such as: when you do streamer.close(), either it will return normally, in which case the data was written to caches, or it will throw an exception, in which case only an unknown subset of data were written, and you need to repeat the operation.
I'm not sure that "Failed to find DHT update future for deferred update response" may cause subsequent node failure. Can you provide complete logs from all server nodes in the cluster? Regards, -- Ilya Kasnacheev пт, 5 мар. 2021 г. в 16:16, 38797715 <[email protected]>: > Hi Ilya, > > It's not easy to build a reproducible environment, which is probably a > problem in use, not necessarily a bug. > > In particular, I want to know the behavior of DataStreamer in case of node > failure and whether it has the function of fail over. > 在 2021/3/5 下午8:16, Ilya Kasnacheev 写道: > > Hello! > > Do you happen to have a reproducer for this issue? I've not seen anything > similar. > > Regards, > -- > Ilya Kasnacheev > > > ср, 3 мар. 2021 г. в 12:31, 38797715 <[email protected]>: > >> Hi team, >> >> When using DataStreamer to write a large amount of data at high speed, if >> one server node fails, then other nodes will appear a lot of following >> information, and finally they will also fail: >> >> [2021-03-03 15:49:44,516][WARN ][sys-stripe-6-#7][atomic] Failed to find >> DHT update future for deferred update response [futId=142939814, >> nodeId=bfe0d1a9-8e0c-4a9e-8b62-041f2f252a80, >> res=GridDhtAtomicDeferredUpdateResponse [futIds=GridLongList [idx=256, >> arr=[142939814,142731432,142827914,142939816,142731434,142939818,142939820,142939822,142939824,142939826,142939828,142731436,142939830,142731438,142939832,142939834,142731440,142827916,142731442,142827918,142827920,142939836,142827922,142939838,142827924,142827926,142939840,142939842,142827928,142731444,142939844,142827930,142731446,142939846,142827932,142939848,142731448,142827934,142939850,142827936,142939852,142827938,142827940,142939854,142827942,142939856,142939858,142827944,142827946,142939860,142827948,142939862,142827950,142731450,142939864,142827952,142731452,142939866,142939868,142827954,142731454,142827956,142731456,142827958,142939870,142827960,142731458,142939872,142827962,142939874,142731460,142939876,142939878,142731462,142731464,142939880,142731466,142939882,142939884,142731468,142939886,142731470,142939888,142731472,142731474,142939890,142731476,142939892,142731478,142939894,142939896,142731480,142731482,142731484,142731486,142731488,142731490,142731492,142731494,142939898,142939900,142731496,142731498,142939902,142939904,142731500,142939906,142939908,142731502,142939910,142731504,142939912,142939914,142731506,142939916,142731508,142731510,142939918,142939920,142939922,142731512,142939924,142731514,142939926,142731516,142939928,142731518,142939930,142731520,142939932,142939934,142939936,142731522,142939938,142731524,142939940,142731526,142939942,142939944,142731528,142939946,142731530,142939948,142939950,142731532,142939952,142939954,142731534,142939956,142731536,142939958,142731538,142939960,142731540,142731542,142939962,142939964,142731544,142731546,142731548,142939966,142731550,142939968,142731552,142939970,142731554,142731556,142939972,142731558,142731560,142731562,142939974,142731564,142939976,142939978,142939980, >> >> What I want to ask is: >> >> 1.Are these logs related to node failures when the DataStreamer writes? >> >> 2.Does DataStreamer have a mechanism for fail over? We know that >> DataStreamer sends data to specified nodes in batches. When a node >> fails, what is the behavior of DataStreamer? >> >
